Speaker Impedance Switch combined with Negative Feedback Switch (lots of states)

[scstill]

I have included a OT impedance switch on many builds to add flexibility to the speaker choice.
A current 5F1 build uses negative feedback with 22k resistor that I wanted to switch out for more distortion.

Rob talks of a 3way switch that includes a heavy feedback option for extra clean I guess https://robrobinette.com/Generic_Tube_Amp_Mods.htm
But to do this with a switched impedance would require 12 states on this NegFB 3way
I think I can accomplish this with the 12pin on/off/on switch in my back of the envelop drawing.
I wanted to get your positive or negative feedback (haha) on the design as well as suggestions on the resistor values?

[shooter]

I would use the search feature in this section for NFB
I very seldom use NFB, but my recollection NFB is "tied" to a speaker tap NOT speaker impedance.
Your switch definitely violates the KISS principal.  any fail, mechanical, design, could leave the PA without a load, making it really easy to kill tubes at will.

[sluckey]

Use separate switches for NFB and speaker impedance. The NFB should be connected to one OT tap only. Don't switch taps! Just use the NFB switch to select different value feedback resistors. And use the impedance selector to switch OT secondary taps to the speaker jack.

[scstill]

This attachment by Rob is what led me to believe that I should use different OT taps and that NFB value changes by impedance. Why would my design not follow this approach?

BTW - it does use two switches, one for impedance, one for NFB

BTW2 - I tried different  speaker impedances with the 22k value and could not tell the difference
which might speak to the KISS concept.

[sluckey]

Of course the amount of NFB will change if you change OT secondary taps without changing anything else. That's because the voltage will be greater on a higher Z tap. I'm surprised you could not hear a difference when changing speaker taps. Maybe the 5F1 is not a good test bed for this. I think you may have missed the point of Rob's speaker tap/NFB chart. Maybe spend some more time on it until you understand what he was saying.

My point about using two switches was based on the requirements for the switches. The speaker switch needs to be heavy duty to handle high current while the NFB switch can be a whimpy low current switch since NFB is a voltage function that doesn't require high currents. It's unlikely you can even find an affordable switch with low current contacts as well as high current contacts, so you would need to look for a complicated switch that has all high current contacts.  Good luck with that.

[scstill]

I double checked and believe that at the 350v primary the most I will see on secondary is 2a with the 4ohm tap
I am using the Carling switch below for the speaker switch.

[scstill]

The text from Rob (below) also seems to imply a change in NFB value when changing the OT tap, which is what I used to determine the values for different impedance selection.

"When changing the feedback source from one output transformer secondary to another you change the feedback resistance by a factor of 1.41 for one step change (for example going from a 2 ohm speaker tap to 4 ohm), a factor of 2 for 2 steps (example 2 ohm to 8 ohm), or a factor of 2.83 for 3 steps (example 2 ohm to 16 ohm)."

[scstill]

I'm surprised you could not hear a difference when changing speaker taps. Maybe the 5F1 is not a good test bed for this.

In one case I used an 8ohm 8" and in the other case a 4ohm 2x12. So that might have had something to do with it :-)

[sluckey]

Aha! You don't have to change the speaker tap when you change the NFB tap. You change the speaker tap to make the output tube happy. You change the NFB tap to make your ears happy.

[kagliostro]

I'll connect the NFB to the tap with the higher voltage (Z) and use a pot to vary the NFB that go back to the circuit


Franco

[scstill]

For the 5F1 build, I have tried many NFB resistance combinations 500 to 100K, also the pot and NC. I cannot really decern much of a difference, if at all.
Might be my ears, although seems that there should be some difference.
Tried different guitars, different 5F1 volume, clean to full drive. Same result.
Rob Rob says best mod ever so I had to try.
The circuit in the diagram, if anything jumps out please let me know. TY

[sluckey]

I wonder if the NFB is phased properly. Swap the primary leads and see what happens.

[scstill]

swapped OT primaries blu to  PS filter, red to 6v6 plate
no change with NFB 22k or NC

[sluckey]

That's crazy. Are you sure the NFB connection to the OT is solid? No broken wire?

Try this. Replace the 22K with a 2.2K. This should give a LOT of NFB. Surely you can hear a difference now.

[scstill]

I tried on another 5F1 build with same results.
But have determined there are several dynamics going on that I just discovered...

First, Fenders original 5F1 used 22k with a 4ohm speaker.
I'm using an 8ohm, so I should adjust the NFB based on the higher impedance.
According to Rob this should increase by a factor of 1.4 so should use 30ohm to get original 5F1 design (I think).

Second, I have a MV preceding the 6v6 which I previously had been running at half.
When I max the MV (how a 5F1 would run) you clearly get a lot more crunch with the NFB out of circuit. Of coarse now my ears are ringing with the vol set to half. :-)
BTW - 2.2k really sucks the life out of it :-(

Still don't know how guys can play the 5F1 without MV in the house, anything above vol of 3 is deafening. Maybe they turn guitar down. I like that I can run high vol for some crunch and still have sanity with MV down. I have used power soaks and Lpads. In fact have LPad before the speaker on my Pro Juniors to get the crunch. Coarse you can always plug in the pedal board.

[sluckey]

BTW - 2.2k really sucks the life out of it :-(

That was the point. At least now you know the NFB is working.    I didn't suggest that to be a proper fix. I just wanted to see if you could finally hear a difference. If you had not heard it suck that would have meant there was a break in the NFB path .

[scstill]

This is where I landed. Not so KISS, but good learning and layout experience.
One switch for impedance (make tubes happy)
One switch for NFB (Make ears happy)
Sounds good...I decided to go with:
"Normal Clean" 22k NFB @ 4ohm (fender 5F1 design reference) scaling to 8 and 16ohm (1.4 times each step)
"Extra Clean" reducing by 3 from normal (best sound for me)
"Dirty" with no NFB.
 

[PRR]

What's done is done, and looks good. But next time....

The three taps are 99.9 equivalent except for voltage ratio. Which you will be adjusting in your resistors. For all practical purpose you arbitrarily pick ONE tap, figure the NFB for that, and it will work on the other output taps also.

Look at almost any Hi-Fi amp. They don't switch NFB connections or resistors.

Here you want "lower fi" NFB choices, so you do need several throws. But all on one pole.

[scstill]

I noticed that as well.
When I was testing with other speakers the sound was comparable with a single set of resistors.
I decided to give each its own NFB value based on that I read at Robs site for NFB with varying impedances (earlier post)
But what's done is done. :-) For me it's always a fun creative challenge to find best point to point layout.
BTW - this 12pin switch (used as 9pin) was about the same price as a 3 pin switch and there are hundreds of resistors around here so not much of a cost. I will probably just play it with normal which sounds best and rarely will change to another speaker. Thinking that in the future an NFB on or off might be the way to go....